Mobile network operators around the world are facing an unprecedented wave of collective anxiety. On the one hand, users’ demands for network speed and stability are constantly rising. On the other hand, traditional traffic-based billing models are gradually reaching their limits.
Faced with a steady drop in average revenue per user, the telecom industry is reaching a consensus. It must closely combine cloud-native systems with artificial intelligence. This is also the only way out of the profitability crisis.

I. No Other Choice: Why Must Mobile Networks Go “Cloud-Native”?

“The entire industry is irreversibly moving toward cloud-native—we have no other choice.” This is the candid assessment of a senior telecommunications network expert.

In the past, traditional mobile networks relied on a large amount of proprietary hardware. Cloud-native technology, however, software-defines network functions. This means networks can operate just like internet applications—enabling expansion within seconds, automatic recovery, and rapid iteration.

However, the transition has not been smooth sailing. Many operators have found that moving from traditional networks to standalone 5G networks is much slower than expected. In the early stages, adopting a cloud-native architecture meant operators shifted focus. They moved from developing new services to maintaining complex cloud-based networks.

This situation, where “efforts are tied up in infrastructure,” calls for a better tool to break the deadlock.

II. Solving the Profitability Puzzle: New Business Possibilities Brought by AI and Open APIs

In the second half of the 5G evolution, how can network traffic be converted into real revenue? The industry is exploring the following highly promising directions for the integration of AI and networks:

AI-Empowered Next-Generation Voice and Interaction:

Enabling traditional voice calls with real-time translation, smart assistant capabilities, and multimodal interaction.

Monetizing Network Data Assets

Leveraging underlying network data streams to provide precise predictive and analytical services for vertical industries such as transportation and logistics, while ensuring privacy.

Commercializing Authorization and Identity Authentication: 

Capitalizing on operators’ inherent “trust attribute” to provide more secure and convenient digital identity verification for the financial and e-commerce sectors.

Opening APIs for AI Agents

This is currently the most closely watched area of innovation. Through the “Network-as-Code” concept, telecom operators can package complex network capabilities into standardized APIs.

In the future, third-party developers and system integrators will build AI agents. They will use APIs to request the best network bandwidth, latency, and slicing services. The creation of this ecosystem is expected to help telecom operators move beyond being “pipeline providers.” It should position them as “core enablers of the AI ecosystem.”

However, this revolutionary transformation also comes with challenges. Virtually all innovative business models currently face complex cross-border compliance and regulatory constraints.

III. Soaring Hardware Costs and the Birth of the “AI Grid”

Behind AI’s big vision, the industry faces a grim reality. Server and hardware costs keep rising, driven by the global AI infrastructure boom.

Because computing resources are scarce and expensive, operators are locked in a tug-of-war with suppliers. They argue over hardware depreciation and longer lifecycles. To address this challenge, several technological frontiers are undergoing changes:

A More Open Approach to Public Clouds

To alleviate the pressure of building their own data centers. Some operators are adopting a more open attitude toward using public clouds to host core network functions.

Efficient Reuse of Hardware

The industry is exploring ways to integrate high-performance general-purpose processors and GPUs into standard edge servers.

This convergence of software and hardware has given rise to the concept of the “AI mesh.” Future edge network nodes will not only be responsible for forwarding mobile signals. But their built-in GPUs will also be capable of processing thousands of tokens per second. This means providing local computing power directly at the network edge to support nearby AI applications.
 

IV. “Should I Hand Over the Keys to the Network to an Artificial Intelligence Agent?”

As AI becomes part of the network, a key question arises. Can we allow AI to manage the nation’s communications lifeline on its own?

Experts offer this pragmatic advice: “Let AI agents handle the heavy lifting.”

At this stage, telecom operators do not need to directly relinquish ultimate control. Instead, they should let AI handle difficult operations and maintenance work. AI can collect network logs, it can also find the root cause automatically.

Through this “human-machine collaboration” automation strategy, telecom operators can free employees from tedious infrastructure tasks. They can then focus on developing high-value new services.
 

Conclusion: The Best Is Yet to Come

5G is not merely a technological upgrade; it is a complete transformation of business models. Although the current path to profitability is still unclear, new 5G evolution technologies are emerging. AI is also being used in automated operations and maintenance, as well as API monetization.

Because of this, the most exciting chapter in the mobile communications industry may be just beginning. For operators facing this maze of unknowns, AI is no longer optional. It has become a turning point on the path ahead.